For all the talk of a trade war breaking out between the U.S. and European Union over online privacy, it may well be that consumers in the two markets have pretty similar expectations regarding what happens with their personal data.
That at least appears to be one of the conclusions drawn by London-based analyst house Ovum, which has just published the results of its annual Consumer Insights survey. The survey, which covered more than 11,000 respondents in 11 countries around the world, included several questions about the use of personal data:
- Are you aware that your personal data is being collected by internet services? Average response: 75 percent “yes”.
- Do you believe internet companies are honest about how they exploit your personal data from tracking? Average response: 14 percent “yes”.
- How true do you believe the following statement is? “Many internet companies are reliant on selling your personal data to advertisers in order to survive.” An average of 46 percent said this was “definitely true” or “true for most companies”.
- Would you select a do-not-track [DNT] feature if it was easily available, when using a search engine? Average response: 68 percent “yes”.
That last question is the kicker, and we’ll get to the implications in a moment. But first, I asked Ovum analyst Mark Little for a bit more insight into the geographical breakdown of his results.
“In Europe we had done research before on online trust, in 2010, and Germany had come out as quite a big country in terms of disliking tracking, and also France. This time round France came out as being the most vociferous potential blockers, more so than the UK and even more so than Germany.”
Eighty-one percent of French respondents apparently said they would implement DNT, versus that global average of 68 percent. But Americans were actually pretty close behind, at around 77 percent — that makes them five percentage points more likely to block web-tracking technologies than their British counterparts. As Little put it regarding the overall impression he gained from the privacy-related questions, “if you took the average of the European countries, it’s probably pretty close to America”.
The comparison is interesting, but it’s the conclusions that bite. Now, obviously with a survey like this you’re going to get a slightly skewed impression of what people will do in practice – 81 percent of French people may say they’d enable DNT if they saw it, but the number that actually would click the button would be much lower. That said…
“Given the size of that figure, even if it was half or a third even of that, we’re talking about quite a large number of users becoming invisible in data terms,” Little stressed. “The issue is that big data has taken what I call little data for granted — they’ve assumed that they will be able to continue data fracking using cookies forevermore and that nothing will ever stop them from doing that. I think that view is incorrect.”
Little contrasts what he terms “data-fracking” – a loaded term for sure – with “data-friending”.
“A data-friending strategy would be much more explicit,” he said. “Otherwise the user will eventually tire of this use of their personal data and will move to a different ecosystem for releasing their data. There is already an emerging personal data ecosystem controlled not by the setting of cookies, but by personal data vaults.”
By this Little meant services such as Personal, which lets users store personal data in a privately-controlled container, then share it with other services in a granular and deliberate way. I can’t say I’ve seen such services take off in a big way yet, but Little seems to see big uptake down the line if web services don’t become more “opt-in” in the way they track and exploit their users.
Of course, the reason the likes of Google and Facebook aren’t too keen on this approach is that they want to harvest as much data as possible. And there again, Little thinks the companies are missing out. Why? Because of the quality of their fracked data.
“Big data insights are estimated and historical – they’re based on what they’ve seen you do; they join the dots and correlate to estimate what you will do,” he said. “In a personal data vault, people can put in what they’re going to be interested in, and what they actually are interested in. It’s them that recorded it – it’s more reliable and it’s future-looking. The quality of the dataset is much better.”
In my analysis, that flies in the face of the big data trend, which is all about increasingly educated guesswork, at least in its predictive sense. But if he’s right, and enough people lose patience with being stalked online by advertisers to the extent that they block them and mess up the resulting datasets, then it would be the logical way forward.
(Incidentally, the tension between personalization and privacy will be the subject of a debate on the first day (March 20) of the GigaOM Structure:Data conference in New York.)